As AI data centre infrastructure expands globally, concerns about its environmental impact are growing. A UN-backed study projects a sharp increase in water and electricity consumption.
Data centres powering artificial intelligence (AI) could consume 9.3 trillion litres of water and 945 terawatt-hours (TWh) of electricity annually by 2030, according to a report by the United Nations University Institute for Water, Environment and Health (UNU-INWEH).
The report estimates that AI-driven data centres could account for nearly three per cent of global electricity consumption by the end of the decade. They are also expected to occupy more than 14,500 sq km of land and generate up to 2.5 million tonnes of electronic waste every year.
According to the study, the projected water consumption would be enough to meet the basic annual needs of around 1.3 billion people in Sub-Saharan Africa. The land required for these facilities would be roughly twice the size of Jakarta in Indonesia.
The findings come as governments and technology companies continue to invest in AI infrastructure to support the growing use of generative AI applications. The report noted that global data centres consumed an estimated 415 TWh of electricity in 2025 and could more than double that figure within the next five years.
Investments in AI infrastructure are also driving demand for advanced computing hardware and expanding data-centre capacity worldwide. The trend is expected to support the deployment of more servers and AI-focused processors in the coming years.
Most AI-related energy consumption now comes from inference, or the day-to-day operation of AI models responding to user requests, rather than training the models themselves.
